Java, C#, VM Performance and Benchmarks

[Dave]

In article <dvqb9h$20j6$1 at digitaldaemon.com>, clayasaurus says...
>
>Craig Black wrote:
>> <snip> But try to write a 3D game engine in Java.  I've talked
>> with some of the best game programmers in the world and they say the same 
>> thing:  Java is damn slow.
>> 
>> -Craig 
>> 
>
>Here are some interesting aspects of Java game development.
>
>1) Quake2 in Java achieves ~85% performance of the C version : 
>http://www.bytonic.de/html/benchmarks.html
>

I just couldn't help it... Had to add - because I've seen this before - that
they've apparently undergone a large amount of effort in tuning the Java code,
but have not tuned the C version along side it (because chances are that quite a
few of the algorithmic changes are applicable to the C code as well).

Plus there isn't any details - like how the C version was built. I know of a
project not too long ago where they burned a lot of cabbage on tuning a Java
app. in a few bottleneck areas. There happened to be some overlap with another
(C++) application, and a while later the Java developers said "Ah-Ha!" it's even
*faster* and went and had a beer. The C++ team came in the next morning, added
'-O2' to the make files, removed some instrumentation, rebuilt, and ... some of
the Java ended up being converted to C++.

Don't get me wrong - the Bytonic stuff is cool, and perhaps is a 100% legitimate
comparison, but I get the feeling that type of stuff goes on a lot (with a lot
of things, not just Java).

>
>2) An interesting blog post about programming Java on cell phones : 
>http://www.armadilloaerospace.com/n.x/johnc/Recent%20Updates